Practical AI Newsletter28 October 2026
AI that edits, designs and works while you are away.

AI that edits, designs and works while you are away.

This week, AI tools moved closer to doing finished work: agents that act in the background, and image models that design and edit rather than just generate. Claude Duet opens as a waitlisted research preview for US Pro and Max subscribers. GPT-6.1 Terra is rolling out in ChatGPT Work and Codex on paid plans. Canva’s Craft is free to try in Dream Lab, and Microsoft’s MAI Image 2.7 brings precise photo editing in preview. Moonshot’s Kimi K3.1 Pro costs $2/$10 per million tokens, with open weights due November 4.

Editor's Picks
Editor's Picks5

MAI Image 2.7 debuts above GPT Image 2.5 Flare on Artificial Analysis and #3 on Arena

Microsoft AI released MAI Image 2.7, an image editing model, in preview on October 27, 2026, debuting at No. 3 on Arena's image editing leaderboard.

Expand details

Introducing GPT-6.1 Terra

OpenAI released GPT-6.1 Terra on October 26, 2026, a fast, lower-cost model for everyday work that reads what is on screen, takes actions across apps and writes more naturally, available in ChatGPT Work, Codex and the API.

Expand details

Canva launches Craft, its first in-house image generation model

Canva released Craft, its first in-house image generation model, on October 22, 2026, in Dream Lab for all users and through its public API.

Expand details

Moonshot's Kimi K3.1 Pro flirts with open weights leadership amid paid influencer scandal allegations

Moonshot released Kimi K3.1 Pro, a 3T-class model, in its API and Kimi apps on October 23, 2026, with full weights promised by November 4.

Expand details

Introducing Claude Duet

Anthropic introduced Claude Duet, a new proactive agent app, as a research preview on October 21, 2026.

Expand details
AI Business
AI Business25

OpenAI expands partnership with Oracle to bring its apps to ChatGPT

OpenAI and Oracle announced on October 27, 2026, an expanded partnership to integrate Oracle applications and data into ChatGPT and power new Oracle offerings with OpenAI models.

Expand details

Anthropic files S-1 with the SEC for proposed initial public offering

Anthropic, PBC filed a registration statement on Form S-1 with the U.S. Securities and Exchange Commission on October 26, 2026, for its proposed initial public offering of Class A common stock.

Expand details

Introducing the SpaceX Super Intelligence Partner Network

SpaceX Super Intelligence (SpaceXSI) introduced its official Partner Network on October 22, 2026, for companies building products and services with Grok.

Expand details

Uber brings ChatGPT Work to all employees with OpenAI

Uber announced the company-wide rollout of ChatGPT Work to over 30,000 employees globally, starting October 22, 2026.

Expand details

Donating Agent Skills to the Agentic AI Foundation

Anthropic donated Agent Skills, an open standard for AI agent specialized knowledge and workflows, to the Agentic AI Foundation (AAIF) on October 22, 2026.

Expand details

HCLTech joins the Claude Partner Network as a Global Premier partner

HCLTech joined the Claude Partner Network as a Global Premier partner on October 21, 2026.

Expand details

Physicist Christopher Stubbs joins Anthropic's Long-Term Benefit Trust

Dr. Christopher Stubbs, a physicist and former Dean of Science at Harvard University, joined Anthropic's Long-Term Benefit Trust on October 21, 2026.

Expand details

Perspective

2

Expert opinions and analysis on how AI is reshaping business strategy, work and decision-making.

How we built ChatGPT’s novel ad stack for user trust and relevance

OpenAI published an engineering deep-dive on October 21, 2026, detailing how it built ChatGPT's LLM-powered ad stack to ensure user trust and relevance.

Expand details

Can we trust LLMs to judge whether a research idea is novel?

Allen Institute for AI (AI2) released a new study and benchmark called "Old Ideas, Novel Problems" on October 27, 2026.

Expand details

AI & Society

5

How AI is changing the way we live, work and think.

From classrooms to careers: The future of education in an AI-driven world

Microsoft announced Learning Curves on October 26, 2026, the second book from its AI Economy Institute, led by Juan M. Lavista Ferres and Frank Nagle, collecting research essays on how AI is changing learning and work.

Expand details

ChatGPT’s expanding world of work

OpenAI released OpenAI Signals data on October 26, 2026, exploring 165 work activities across nearly 100 countries, revealing how consumers use ChatGPT for work.

Expand details

We’re helping farmers around the world connect to essential services online.

Google.org announced a $2.5 million investment to help 1 million farmers access AI-powered agricultural services via the World Bank's ONA by the end of 2027.

Expand details

Microsoft and the Public Library Association expand free AI literacy training

Microsoft and the Public Library Association (PLA) announced an initiative on October 22, 2026, to expand free AI literacy and skills training for public library workers and patrons.

Expand details

How Google Beam connects military families globally

Google partnered with the USO to expand Google Beam to 20 global locations, connecting deployed service members and their families.

Expand details

AI & Safety

6

Safety, security and risk in AI systems, and what is being done about it.

Securing our frontier model training environment

OpenAI published an article on October 23, 2026, detailing how it strengthened the security of its frontier model training environments.

Expand details

Claude Duet Safety, Privacy, and Security

Anthropic published details on the safety, privacy, and security features of Claude Duet on October 23, 2026.

Expand details

Democratic governance of frontier AI, part II: Empowering cyber defenders

OpenAI published "Democratic governance of frontier AI, part II: Empowering cyber defenders" on October 21, 2026, outlining a blueprint for strengthening cyber defenses with frontier AI.

Expand details

OpenAI and Poland’s NASK partner to advance AI safety and security

OpenAI and Poland's NASK National Research Institute signed an agreement on October 23, 2026, to collaborate on AI safety, security, and technical research.

Expand details

Patch the Planet: global response coordination as agents uncover critical vulnerabilities

OpenAI announced Patch the Planet on October 21, 2026, after security firm Calif used GPT-5.6 Sol to find one of two critical PHP flaws and Wiz independently found both; OpenAI coordinated the response with PHP maintainers.

Expand details

Broad access through consistent standards

OpenAI published an update on October 21, 2026, detailing how it works with distribution partners to uphold consistent safeguards and clarify expectations in its Usage Policies for broad access to its models.

Expand details

Research & Science

5

How AI is advancing science, medicine and research, and the programms that put it in researchers’ hands.

Can AI diagnose like a team of doctors? New research puts it to the test

Microsoft AI published research on October 22, 2026, demonstrating that its MAI-DxO AI diagnostic system outperforms physicians in diagnostic accuracy and cost-efficiency.

Expand details

Early progress on accelerating drug development with Claude

Anthropic reported early results on October 27, 2026 from drug development partnerships with six organizations, including Generate Biomedicines, Enveda and Bristol Myers Squibb, and from its own muscular dystrophy program.

Expand details

Working with academic communities to build scientific understanding

OpenAI announced new mathematical and computer science discoveries on October 26, 2026, including proofs for the Kakeya conjecture and the refutation of the Exponential Time Hypothesis.

Expand details

Open call for guest investigators in the life sciences

Anthropic announced an open call on October 22, 2026, inviting members of the life sciences community to apply as guest investigators to work on special projects using Claude.

Expand details

With new ideas and human-directed AI, researchers solve a 31-year-old math mystery

Microsoft Research announced on October 21, 2026, that its researchers, using human-directed AI, resolved the 31-year-old Kannan–Lovász–Simonovits (KLS) conjecture.

Expand details
Product Updates
Product Updates16

AI Assistants

3

New capabilities and experience updates across ChatGPT, Claude, Gemini etc.

An easier way to customize how Gemini works for you

Google introduced a centralized "customize" hub in the Gemini app on October 21, 2026, to tailor how Gemini responds and connects with other tools.

Expand details

Take actions on your notebooks from Gemini Notebook directly in the Gemini app

Google announced on October 23, 2026, that Gemini Notebook is now available as a Connected App within the Gemini app, allowing natural language interaction with notebooks.

Expand details

Claude Cowork and chat come together for Enterprise

Anthropic announced the general availability of the unified Claude experience, combining chat and Cowork, for Claude Enterprise plans on October 23, 2026.

Expand details

Productivity & Work

4

Features that help you work faster and get more done in less time.

Introducing artifacts for legal work

Anthropic added a gallery of ready-to-use legal artifacts to Claude for Legal on October 28, 2026, and launched a Legal Ambassadors Program.

Expand details

Create structured documents using just your voice with Docs Live in Google Docs on mobile

Google Docs rolled out Docs Live to eligible Workspace users of its mobile app on Android and iOS on October 21, 2026.

Expand details

Real-time speech translation in Google Meet expands to 63 languages with automatic language detection

Google Meet expanded its real-time speech translation to 63 languages with automatic language detection on October 22, 2026.

Expand details

Google Vault now supports filtered exports for Gemini in Workspace

Google Vault introduced support for eDiscovery filtered exports for Gemini in Workspace conversations on October 22, 2026.

Expand details

Finance & Business

2

Features for managing money, expenses, tax and everyday business finances.

Expanding Claude for Nonprofits

Anthropic released the Claude for Nonprofits plugin on October 26, 2026.

Expand details

Deel launches Pangaea AI models for HR and payroll, betting specialists can cut the cost of answers by 98%

Deel launched Pangaea, a pair of specialized AI models for HR and payroll, and Emet, a business for expert-created training material and AI performance evaluation, on October 21, 2026.

Expand details

Creative Content

3

Features for writing, design, image, video and music creation.

We’re announcing Genie 4, our new world model for real-time interactive simulation.

Google DeepMind announced its research on Genie 4, a new world model for real-time interactive simulation, on October 27, 2026.

Expand details

Beyond static charts: Helping AI create visualizations you can explore

Microsoft Research published details on October 22, 2026, about Flint-interactive, a new language for creating interactive data visualizations with AI agents.

Expand details

Gain more creative control over Slide Decks and Video Overviews in Gemini Notebook

Google updated Gemini Notebook on October 21, 2026, to provide users with more creative control over Slide Decks and Video Overviews through a step-by-step creation flow.

Expand details

Marketing

1

Features for campaigns, content promotion, audience growth and reaching customers more effectively.

ChatGPT Ads expands to 26 new countries as over 100,000 advertisers have used the platform

OpenAI announced that ChatGPT Ads expanded to 26 additional countries, bringing its availability to 89 countries worldwide, on October 21, 2026.

Expand details

Other Updates

3

Noteworthy AI updates that don't fit neatly elsewhere.

OpenAI models are coming to Google Cloud

OpenAI announced a partnership with Google Cloud to integrate its frontier models and Codex workflows into Google Cloud's Model Garden on October 21, 2026.

Expand details

The full Claude Platform, available on Google Cloud

Anthropic announced on October 22, 2026, that Claude Models on Google Cloud (Anthropic managed) is now generally available, providing Google Cloud customers with the full Claude Platform features.

Expand details

Our newest AI Quest: Mapping the human brain

Google Research and the Stanford Accelerator for Learning launched "Polar Peaks," a new AI Quest lesson on October 26, 2026, focused on mapping the human brain.

Expand details
AI in Practice
AI in Practice16

Case Studies

11

What it looked like when a real company put AI to work.

8 examples of Gemini 4 Argon in action

Google published eight examples of Gemini 4 Argon in action on October 27, 2026, showcasing its capabilities in complex tasks.

Expand details

Retell uses OpenAI models to help power 100M monthly calls

Retell announced its voice AI platform uses OpenAI models, including GPT-6 Astra and Codex, to power over 100 million monthly calls as of October 27, 2026.

Expand details

Legora is building the agentic operating system for legal work

OpenAI published a case study on October 28, 2026 on how Legora, the Stockholm-founded legal AI company, uses OpenAI models, ChatGPT Work and Codex both in its product and inside its own teams.

Expand details

How Fanatics and OpenAI streamline the fan experience

Fanatics launched an OpenAI-powered "Shop the Look" feature to streamline the fan experience on October 26, 2026.

Expand details

Cloudflare links vulnerability discovery to defense with OpenAI

Cloudflare announced on October 21, 2026, that it is using OpenAI Daybreak to investigate vulnerabilities and propose fixes, and Codex to accelerate developer productivity and improve code quality.

Expand details

Your next sales teammate should be a Grok Bot

SpaceXSI published on October 27, 2026 how its sales team uses Shared Grok Bots for prospecting, quoting and deal review; the post is by Palmi Angelova, Joe Robertson and Iris Wang.

Expand details

How Check Point accelerates vulnerability research with Daybreak

Check Point announced on October 27, 2026, that it accelerates vulnerability research with OpenAI Daybreak, reducing investigation times from hours to minutes.

Expand details

How Ultrafast cut incident investigation from hours to minutes at OpenAI

OpenAI published a case study on October 26, 2026, detailing how its engineers use Codex Ultrafast to reduce incident investigation times from hours to minutes.

Expand details

Data and AI agents give a lift to TK Elevator service technicians

TK Elevator service technicians are now using an agentic AI app, created with Microsoft Foundry and Azure OpenAI, to streamline maintenance tasks.

Expand details

Using thousands of AI agents to uncover new leads in biological data

OpenAI, PerturbAI and Scripps Research published on October 21, 2026 how thousands of GPT-Rosalind agents analysed data from millions of mouse brain cells in 32 minutes.

Expand details

How AI could help scientists track a pest threatening citrus trees

Microsoft Discovery app was used in a research project, published on October 26, 2026, to rapidly adapt biological monitoring systems for tracking the Asian citrus psyllid (ACP).

Expand details

Frameworks and Guides

5

How-to advice and frameworks you can apply yourself.

How to track and control team token spend

Perplexity published a guide on how teams can track and control their AI token spend on October 21, 2026.

Expand details

What Is RAG Architecture? A Full Definition With Detailed Practical Examples

Perplexity published a guide titled "What Is RAG Architecture? A Full Definition With Detailed Practical Examples" on October 26, 2026.

Expand details

How to evaluate AI agents: test sets, graders, and regression checks

Perplexity published a six-step guide on October 22, 2026, for evaluating AI agents.

Expand details

Building agents: When to fold subagents into your main agent loop

Anthropic published an article on October 22, 2026, advising developers building agents on the Claude Platform on when and how to integrate subagents into the main agent loop.

Expand details

15 college students share their favorite Gemini features

Google published an article on October 26, 2026, detailing 15 college students' favorite Gemini features for learning and studying.

Expand details
AI & Developer
AI & Developer14

When AI rewrites millions of lines of code, verification becomes essential

Microsoft published an article on October 26, 2026, discussing the challenges and solutions for verifying AI-rewritten code in large-scale agentic software engineering projects.

Expand details

GPT-6.1 Sol helps CodeRabbit strengthen security across complex codebases

CodeRabbit announced its use of OpenAI's GPT-6.1 Sol to enhance security analysis across complex codebases on October 26, 2026.

Expand details

Greptile takes bug detection beyond code review with GPT-6 Astra

Greptile announced its use of OpenAI's GPT-6 Astra, Daybreak, and Codex to enhance bug detection and code review on October 26, 2026.

Expand details

A coding agent’s harness can triple its cost without making it more accurate, study finds

VentureBeat reported on October 23, 2026 a Shandong University study finding that an AI coding agent's harness can triple its cost without improving accuracy.

Expand details

See what 5 builders are making with Gemini 4 Argon

Google showcased five community projects built with its Gemini 4 Argon model on October 23, 2026.

Expand details

Claude Security now scans continuously and tracks fixes

Anthropic updated Claude Security on October 21, 2026, to continuously scan code repositories and track fixes.

Expand details

How does an LLM gateway work? A practical guide for engineering teams

Perplexity published a guide on October 27, 2026 explaining how LLM gateways route, secure and log AI model calls.

Expand details

MAI Cyber 1.1 Flash: Sharper findings and fewer tokens inside MDASH

Microsoft AI released MAI Cyber 1.1 Flash on October 27, 2026, a cybersecurity model inside its MDASH vulnerability scanning system.

Expand details

Manage costs and demonstrate ROI for Claude Code on Google Cloud

Anthropic published a guide on October 27, 2026 on controlling spend and measuring return on investment for Claude Code on Google Cloud.

Expand details

Build specialized multimodal models with LFM3

Liquid AI released LFM3, a new generation of Liquid Foundation Models, on October 27, 2026.

Expand details

Does Your Model Run on AWS Trainium? Look It Up

Hugging Face launched a "torch-neuronx model support" Space on October 23, 2026, which allows users to check if their Transformers models run on AWS Trainium.

Expand details

LitPivot: Developing research ideas through dynamic contextualization within the literature landscape

Allen Institute for AI (AI2) developed LitPivot, a system designed to help researchers develop and refine research ideas by dynamically contextualizing them within existing literature, which will be presented at UIST 2026.

Expand details

Rapidly scaling online storage to serve over 1 billion ChatGPT users (Part II)

OpenAI published Part II of its series on October 23, 2026, detailing how it scaled its Habitat online storage platform using Azure Cosmos DB to serve over 1 billion ChatGPT users.

Expand details

Build more responsive apps with the Decisions API

OpenAI made its Decisions API generally available on October 22, 2026.

Expand details
Practical AI Newsletter

Join our free weekly newsletter. We select, explain and categorise AI news so you know what matters, what to watch out for and how it can be used.

Powered by Sender.net. Unsubscribe anytime.